UAs, Barnacles and other mysteries Thread 8 - The Canonn

Status
Thread Closed: Not open for further replies.
Well guys, ive come to the conclusion that theres a wreckage/strandage on the dark side of M5C. Im 8500LY/8 straight hours of flight out on my way back from a multi month exploration expedition/Jaques and plan on stopping by Gateway to drop off my data, then Im going to go look for the wreckage. If i start after work i should be there around 2-3am central time. Im Demon Eyes Cain on XBL if any CMDRs would like to join me. Wish me luck, ill post here and on reddit if i find anything. I feel like ive got a good shot because im unusually lucky when it comes to stuff.

Good lucl but bear in mind that 5c doesn't have a dark side.
 
Soo I finally have arrived at Jaques Station and since FD added new sounds to the game yesterday I've checked for some odd ones. Yesterday I've some posts on the forum about weird patterns in station sounds, primally from the comms arrarys in the station inside (The long antenna thingy in front of you when you enter a station.. with the red blinking light.). I started audacity did live spectrum analyser and went inside, guess what I've found.. weird shirps and odd patterns in my spectrum! :eek: I listened to it for about 5 minutes and that I had the feeling that the patterns are repeating and went outside and stopped the analyser.

I've heard that UA infected station only have this patterns and I guess Jaques Station was also infected by the UAs. If I'm wrong it's weird and I guess that would also be a new discovery.

Here is the the audio as a spectrum, I had to make 8 pictures because it was to long for one.


I'm super sorry if this has already been posted on this Thread, I'm not up to date anymore.
 
Last edited:
Morse code is a set of dots/dashes, of variable length. (v is 4 bits, e is 1).

Listening to a portion of the UP audio, it instead seems to consist of 'blocks' of 3 bits.

|-- ||- |-- ||- -|- etc.

Which seems similar to the binary on the outside of the image, rather than morse code.

You mean those beeps in low and high tones that come in each cycle of howls, etc? If so, I've been thinking the same thing.

That particular little part of the sound has come up and gone again in discussion many times over, but we never had a sonograph with 3bit symbols in it before.
 
the input frequency.

e.g. peaks at 500hz, 5khz and 15000khz are plotted at log(500), log(5000), log(15000) then scaled to the output height of the bitnap into which the spectrogram is drawn

That doesn't tell me what the logarithmic scaling factor is.

This is an example with a log scale of 10 where each increment in the the y axis height is 10 times the previous value.
linear_log.jpg


The point is, the log factor doesn't have to be 10.
 
Last edited:
You mean those beeps in low and high tones that come in each cycle of howls, etc? If so, I've been thinking the same thing.

That particular little part of the sound has come up and gone again in discussion many times over, but we never had a sonograph with 3bit symbols in it before.

Is it my dodgy memory or didn't the UP visual flashes come in groups of three too? Must look at the video...
 
Morse code is a set of dots/dashes, of variable length. (v is 4 bits, e is 1).

Listening to a portion of the UP audio, it instead seems to consist of 'blocks' of 3 bits.

|-- ||- |-- ||- -|- etc.

Which seems similar to the binary on the outside of the image, rather than morse code.

That's a thing that is rolling in my mind for a long time as well:

we could decode the hi low PURRS, using the 4 symbols on the sonogram. The thing is we only have 4 symbols, while the purrs triplets come in more then 4 variants.

NetSlayer transcribed all the sequences here:
https://docs.google.com/spreadsheets/d/1Xo72JTMlXu8BR49SH_NyXBuQ0dVfg9ifrA7u_k8BP0c/edit#gid=0

but I suggest to wait for a new clean 6 minutes recording of the UP, because that one is a one hour recording, made on some planet's surface, where the UP does not degrade.

Other problem is that, if the symbols are really binary numbers, we could have already decoded the purrs, without the sonogram.
If, instead, each symbol of the sonogram represent something else, related to the four quadrants, then we could decode not a sequence of random numbers, but a sequence of things/actions/angles/whatever, depending on the meaning of the four quadrants.

All depends on the meaning we give to the four quadrants, represented by the four sumbols (-|-, -||, --|, |--).

So, first thing to do after the new UP recording, is to transcribe all the HI LOW PURRS triplets, and then try to "decode" THEM using the symbols on the sonogram.
 
Last edited:
Just an idea that popped in my mind if the sphere is our bubble and the 2 lines in the lower left quadrant are pointing to the Barnard's loop and the permit restricted bubble? Maybe the UP is a key for the restricted zone...

After 4 days searching for convoys in ross 47 without success i'm heading for the COL 137 restricted region in search for possibile free floaters...

Wish me luck!

If anyone with an UP wants to try I'm up to cooperate if it opens the way
 
Last edited:
Guys, someone linked me a recording of what i think is the recording of the UP.

https://soundcloud.com/user-536207472/voice

If you just listen, you can hear at the start something like "We are, we are, we are"

Is it just imagination. Has anyone actualy tried listening to the sound at different speeds, or like you know, hidden messages in rock LPs that when played backwards told people to worship the devil... :p

Has anyone tried listening to a reversed recording? Has anyone tried reversing the recording and then sampling it?

We're preprogrammed to hear voices in noise... Just like we are recognizable objects in random shapes... I think it's just a repeating pattern...
 
Quick questions on the scaling thing.

Do the audio tools let you pick a scale or is it either linear or logarithmic (10)?
I think that the central image is meant to be an exact circle, so is it an exact circle at log (10) or not?

There are two effects going on (mentioned in a post of my earlier this morning) that impact the shape of the 'circular' object.

  1. Scaling factor for frequency. When the image was initially encoded some scaling factor to translate Y axis values into frequency was used, clearly something logarithmic in nature, that could have been simple log or more complex logarithmic scalings like MEL or Bark. MEL would be the most common. (In other words this defines how frequencies are binned into a pixel)
  2. Time vs frequency relationship. This is a spatial relationship. In the original drawing pre-encoding a second represented some number of X pixels across some distance (say an inch), the same number of pixels at the same spacing represented some range of frequencies. During encoding this relationship is there at a low level in the nature of the proper FFT window size and shape. There's no right answer a priori but once the encoding is done there is, it's baked into the audio and the nature of time vs frequency. (This defines the height of a pixel, but says nothing about the frequencies represented in them)

Now when somebody reversed the process they have to select the right scaling factor, ie log, MEL, Bark or whatever. This is easily seen in linear mode where the shape is decidedly deformed, it's less obvious when comparing log, mel or bark. This controls which frequencies get binned together.

They also have to set the proper time vs frequency relationship. This sets the height of a pixel. I suspect by default Audacity set the X and Y spacings as a function of the FFT windows, basically reversing (2) above but it's possible they don't and just make it fit the existing window. Even if Audacity does do get it right initially the person can zoom both X and Y independently in discrete steps and many folks are probably zooming in X and then stretching in Y to make it a circle again.

In terms of whether it should be considered an ellipse vs circle you can't actually tell visually. There's no 'alignment' marker. If you take a circle and stretch it in one axis you get an ellipse. It's not possible to say which is correct after the fact. Taking an ellipse and compressing it till it's a circle is no more correct than taking a circle and stretching till it's some ellipse. You have to figure out the proper relationship from (2) in a mathematical sense, I havent' thought about it in detail but it should be possible to determine what the proper relationship is (it's basically uniform pixels in X and Y over some distance), ensure the display has that and then you can definitively say 'circle' or 'ellipse'.

Personally I suspect it should be a circle because there's only one valid circle but an infinite number of valid elipses. But if somebody wants to definitively state it's a circle or an ellipse then they have to show mathematically what the proper relationship is in (2) above. Otherwise it's just opinion masquerading as fact. This is even more of an issue in the ellipse case, a circle is interesting by it's mere existence, an ellipse is interesting in it's major/minor axis relationship and so you have to know *exactly* what the stretch factor is, just stretching it till it's an ellipse and saying 'AHA' conveys zero information.
 
Last edited:
I thought the 4 quadrants were just 1,2,3,4 but depending on the endianess, each quarter is a clue to the key and there is other info in there we need to decode on order of the quadrants.

But hey, this is all way beyond me. I would never even have had a clue how to decode the sound into an image in the first place.

To be honest, it's quite annoying for me that either this mystery requires genuis levels brainpower to solve, or its going to be something so simple we facepalm when we find out what it is. Whatever, it basically excludes a large part of the community who simply don't understand this stuff and perhaps even more so, those who have little interest in following the forums, and miss out on all this out-of-game stuff. Ideally, this sort of thing should actually be solvable in game.

But i think whoever does solve this, will deserve a PhD in Brooksoligy and a Diploma in Thinking Outside the Box.
 
That's a thing that is rolling in my mind for a long time as well:

we could decode the hi low PURRS, using the 4 symbols on the sonogram. The thing is we only have 4 symbols, while the purrs triplets come in more then 4 variants.

NetSlayer transcribed all the sequences here:
https://docs.google.com/spreadsheets/d/1Xo72JTMlXu8BR49SH_NyXBuQ0dVfg9ifrA7u_k8BP0c/edit#gid=0

but I suggest to wait for a new clean 6 minutes recording of the UP, because that one is a one hour recording, made on some planet's surface, where the UP does not degrade.

Other problem is that, if the symbols are really binary numbers, we could have already decoded the purrs, without the sonogram.
If, instead, each symbol of the sonogram represent something else, related to the four quadrants, then we could decode not a sequence of random numbers, but a sequence of things/actions/angles/whatever, depending on the meaning of the four quadrants.

All depends on the meaning we give to the four quadrants, represented by the four sumbols (-|-, -||, --|, |--).

So, first thing to do after the new UP recording, is to transcribe all the HI LOW PURRS triplets, and then try to "decode" THEM using the symbols on the sonogram.

I think the beeps need transcribing. They come in discreet packets in each cycle, different each time. Maybe the sonograph is a key to where to find the start of the message by finding a sequence of 4 sets of beeps that matches the 4 symbols (I know, it all depends how we read the picture)
 
Ok, I've done a little bit of research.
Audacity is using 10 as it's scaling factor for the logarithmic scale.

The MEL and Bark scales are based on subjective measurments so I don't think they are appropriate.

So, I'll come back to my original question. Using the log10 scale, is it a circle or not?
 
Ok, I've done a little bit of research.
Audacity is using 10 as it's scaling factor for the logarithmic scale.

The MEL and Bark scales are based on subjective measurments so I don't think they are appropriate.

So, I'll come back to my original question. Using the log10 scale, is it a circle or not?

The base of the logarithm doesn't actually have an impact on the image. No matter what base you choose, you still get two axes where the X axis is in units of seconds, and the Y axis in in units of hertz. They're different units, so there's no "correct" way to line them up so your x pixels match your y pixels. It's completely the choice of whoever is doing the decoding. The ONLY real distinction is if you use linear - with any scaling - you get a family of pear-shaped symbols. If you use logarithmic plotting - again, with any choice of scale factors - you get a family of ellipses. One choice of scale factor makes the ellipse a circle, so that's presumably the intended one.
 
I thought the 4 quadrants were just 1,2,3,4 but depending on the endianess, each quarter is a clue to the key and there is other info in there we need to decode on order of the quadrants.

But hey, this is all way beyond me. I would never even have had a clue how to decode the sound into an image in the first place.

To be honest, it's quite annoying for me that either this mystery requires genuis levels brainpower to solve, or its going to be something so simple we facepalm when we find out what it is. Whatever, it basically excludes a large part of the community who simply don't understand this stuff and perhaps even more so, those who have little interest in following the forums, and miss out on all this out-of-game stuff. Ideally, this sort of thing should actually be solvable in game.

But i think whoever does solve this, will deserve a PhD in Brooksoligy and a Diploma in Thinking Outside the Box.

I can sympathize with this. I think this whole thing is really cool, but without any progress I'm kind of useless. I don't really have the willpower or knowledge to crack some puzzle that appears to be either really hard, really easy, or not solvable at the moment. My participation is basically reading the thread, having some fun quips, and checking out Merope and looking at points of interest.

Now the second we have firm knowledge that "Hey we need to go check out THIS system, or THIS nebula because it seems awfully likely that's where something significant is hiding" i'll be ready to go. Right now it's just a waiting game and I feel a bit helpless.
 
The base of the logarithm doesn't actually have an impact on the image. No matter what base you choose, you still get two axes where the X axis is in units of seconds, and the Y axis in in units of hertz. They're different units, so there's no "correct" way to line them up so your x pixels match your y pixels. It's completely the choice of whoever is doing the decoding. The ONLY real distinction is if you use linear - with any scaling - you get a family of pear-shaped symbols. If you use logarithmic plotting - again, with any choice of scale factors - you get a family of ellipses. One choice of scale factor makes the ellipse a circle, so that's presumably the intended one.

It makes a massive difference!!

For example, of the factor was 2, the y-axis would read 2,4,8,16,32 instead of 10,100,1000,10000,100000.
The image would look totally different.
 
Try doing it in perspective. I think.

Big respect for the work either way, though - +rep!

Well, Solidworks (my CAD software) can't (or I don't know how) do drawing views in perspective. I CAN view the 3D model in perspective, so I tilted it 5 degrees below the equator plane (5 deg is a default increment in Solidworks, no other reason for that value) and took a screenshot. Then I did the same analysis as the spectograph image.

JSikS8C.png


Since I overlaid a screenshot image onto a plane, then scaled it in solidworks, I had to manually position the intersection points. Hence, this time it is expected there won't be perfect round values for the angles.

Result:

WBlOF7m.png


Observation: My results are still within a fraction of a degree of accuracy. Granted the lines aren't as fuzzy in the screenshot of my model, which makes the points easier to position. But the spectograph version of this method produced very different angles that continued to decrease as I strayed from the grid center in the Y axis (and even the X axis, although the X axis difference is symmetrical, just needs stretching).
 
Status
Thread Closed: Not open for further replies.
Back
Top Bottom